MHW Dashboard
Nima Farchadi
4/10/2022
Marine Heatwaves Impacts on Fisheries
Marine heatwaves (MHWs) are one aspect of climate change that are already having a significant impact on fisheries. MHWs are short-term extreme warming events that occur globally often linked to disruptive changes on ecosystem services, such as fisheries, with extensive socioeconomic consequences.
#loading in northern region data
knitr::include_graphics("oliver_MHW_graphic.png")
Figure from Oliver et al. 2021
How MHWs Are Impacting U.S. Fisheries?
The anomalous warmer water temperatures associated with MHWs can create physiological stressful environments for marine species if thermal thresholds are exceeded, leading to a wide range of impacts on marine ecosystems such as mass mortalities and strandings, shifts in species distributions, and altered ecosystem structure and functioning.
Regional case studies have shown that these ecological responses to MHWs are linked to pervasive socioeconomic impacts on fisheries, such as shifts in fishing grounds and reduced fisheries yields or distribution changes for species of conservation concern.
Insert infographic
# MAB_NEC_NED_MHW<-here("data","water_temp","NWA","oisst",
# "MAB_NEC_NED_MHW.rds") %>% readRDS()
#
# MHW_contour<-MAB_NEC_NED_MHW %>% filter(year == 2012) %>%
# group_by(lat,lon) %>%
# summarise(temp_anomaly = mean(temp_anomaly, na.rm=TRUE),
# seas = mean(seas, na.rm=TRUE),
# .groups = "drop") %>%
# mutate(MHW = if_else(temp_anomaly >= seas, 1, 0)) %>%
# # sf::st_as_sf(coords = c("lon","lat"), crs = "+proj=longlat +ellps=WGS84 +datum=WGS84 +no_defs +towgs84=0,0,0") %>%
# # mutate(lon = sf::st_coordinates(.)[,1],
# # lat = sf::st_coordinates(.)[,2]) %>%
# # st_union() %>%
# # st_buffer(1000) %>% st_transform(crs = "+proj=longlat +ellps=WGS84 +datum=WGS84 +no_defs +towgs84=0,0,0")
# dplyr::select(lon,lat,MHW) %>%
# rasterFromXYZ() %>%
# rasterToContour() %>%
# st_as_sf()
#
# st_crs(MHW_contour)<-"+proj=longlat +ellps=WGS84 +datum=WGS84 +no_defs +towgs84=0,0,0"
#
# MHW_contour<-st_intersection(world %>% filter(name == "United States"),MHW_contour)
#
# world <- ne_countries(scale = "medium", returnclass = "sf")
#
# ggplot() +
# geom_sf(data = MHW_contour, fill = "red") +
# #geom_tile(aes(x=lon,y=lat, fill = temp_anomaly))+
# #geom_point(NWA_PLL_2019, mapping = aes(x=X, y=Y)) +
# #geom_sf(data = NWA_PLL_zones, color = "goldenrod3", fill=NA)+
# #geom_contour(aes(x=lon,y=lat, z = temp_anomaly),binwidth = 5)+
# geom_sf(data = world, color= "black", fill = NA) +
# scale_size(range = c(1, 8), name=expression("Proportion\nof MHW"(km^2)))+
# coord_sf(xlim = c(-75, -61), ylim = c(35.7, 45), expand = TRUE) +
# ylab("Latitude") + xlab("Longitude") + theme_classic()
# #+scale_fill_viridis(discrete=FALSE, option="turbo",
# #name=stringr::str_wrap("SSTa\nMagnitude (°C)", width = 13))MHWs Come In Many Flavors
Because each MHW is caused by a range of complex oceanographic and atmospheric processes, a single MHW event can have differential impacts on fishing fleets, fishing management areas, and fishing communities.
To better understand how marine heatwaves have already impacted US fisheries, we used shipboard tracking data to map how two US fleets - the pelagic longline fleet in the Northwest Atlantic (NWA) and the trolling fleet in the Northeast Pacific (NEP) - responded to MHWs from 2012-2020.
world <- ne_countries(scale = "medium", returnclass = "sf")
MAB_NEC_NED_MHW<-here("data","water_temp","NWA","oisst",
"MAB_NEC_NED_MHW.rds") %>% readRDS()
NWA_flavors<-MAB_NEC_NED_MHW %>% filter(month %in% c(9,10,11) &
year %in% c(2015,2017,2019,2018)) %>%
mutate(year = factor(year, levels = c("2018","2015",
"2019","2017")))
NWA_PLL<-here("data","AIS_processed","NWA_PLL",
"Pres_Abs_2013to2020_NWA_USA_PLL_onlyfishing_v2_1to1ratio_absenceconstrained_convexhull_v2_enhanced.rds") %>% readRDS()
NWA_PLL_flavors<-NWA_PLL %>% filter(month %in% c(9,10,11) &
year %in% c(2015,2017,2019,2018) &
Pres_abs == 1) %>%
mutate(year = factor(year, levels = c("2018","2015",
"2019","2017")),
Geartype = ifelse(Pres_abs == 1,"Longline","none"))
NWA_flavors %>%
group_by(lat,lon,year) %>%
summarise(temp_anomaly = mean(temp_anomaly, na.rm=TRUE),.groups = "drop") %>%
#mutate(date = as.factor(yearmon)) %>%
ggplot() +
geom_tile(aes(x=lon,y=lat, fill = temp_anomaly))+
geom_point(NWA_PLL_flavors, mapping = aes(x=X, y=Y, color = Geartype),
shape = 1) +
geom_sf(data = world, color= "black", fill = "grey") +
scale_size(range = c(1, 8), name=expression("Proportion\nof MHW"(km^2)))+
coord_sf(xlim = c(-75, -40), ylim = c(35.7, 50), expand = TRUE) +
theme_classic()+
labs(x = expression(weak%<-%Intensity%->%strong),
y = expression(small%<-%Size%->%large))+
facet_wrap(~year)+
scale_fill_viridis(discrete=FALSE, option="turbo",
name=stringr::str_wrap("SSTa Magnitude (°C)", width = 13))+
scale_color_manual(values = "black")+
theme(strip.background = element_blank())
NWA longline fleet responses to varying MHWs during the fall season (Sep, Oct, Nov)
Fishing Area Gains and Losses
To understand how fisheries are responding to MHWs, we took a management zone approach…
Northwest Atlantic Pelagic Longline Fishery
#mangement zones shapefile
NWA_PLL_zones<-here("data","shapefiles","Zones_PLL.shp") %>% sf::st_read(crs = "+proj=longlat +ellps=WGS84 +datum=WGS84 +no_defs +towgs84=0,0,0", quiet = TRUE) %>% st_transform()
#center coordinates for each mgmt_zone
mgmtzone_centroid<-NWA_PLL_zones %>% st_centroid() %>% st_coordinates() %>% as.data.frame() %>% suppressWarnings()
mgmtzone_centroid<-cbind(NWA_PLL_zones$ET_ID,mgmtzone_centroid)
colnames(mgmtzone_centroid)<-c("mgmt_zone","X","Y")
ggplot() +
geom_sf(data = NWA_PLL_zones, color = "goldenrod3", fill=NA)+
geom_sf(data = world, color= "black", fill = "grey")+
geom_label_repel(data = mgmtzone_centroid, aes(x = X, y = Y,
label = mgmt_zone),
fontface = "bold",
box.padding = 3.5,
segment.size = 1,
max.overlaps = Inf,
)+
theme_bw()+
coord_sf(xlim = c(-100, -20), ylim = c(0, 55), expand = TRUE)+
labs(y = "Latitude", x = "Longitude")
U.S. Atlantic Pelagic longline management zones
#Changes in Suitable Fishing Habitat
VDM_stack<-stack(here("data","Monthly_Spatial_Predictions","NWA_PLL","NWA_PLL_monthly_predictions_2012to2020.nc"))%>% suppressMessages()
tt<-as.yearmon(seq(as.Date("2012-01-01"),as.Date("2020-12-01"),by="month"))
VDM_stack<-setZ(VDM_stack,tt)
tt <-tt[tt %in% as.yearmon("Sep 2015")]
#subset to just Sep 2015
VDM_Sep_2015<-subset(VDM_stack, which(getZ(VDM_stack) %in% tt)) %>%
rasterToPoints() %>% as_tibble()
colnames(VDM_Sep_2015)<-c("x","y","suitability")
VDM_Sep_2015<-VDM_Sep_2015 %>% st_as_sf(coords = c("x","y"), crs = "+proj=longlat +ellps=WGS84 +datum=WGS84 +no_defs +towgs84=0,0,0")%>%
mutate(lon = sf::st_coordinates(.)[,1],
lat = sf::st_coordinates(.)[,2])
sf_use_s2(FALSE) %>% suppressMessages()
VDM_Sep_2015 <- st_join(VDM_Sep_2015,NWA_PLL_zones) %>% as_tibble() %>%
filter(ET_ID %in% c("MAB","NEC","NED")) %>% suppressMessages()
#subset to Sep 2015
df<-NWA_flavors %>% filter(year == 2015) %>%
rename("mgmt_zone"="ET_ID") %>%
mutate(temp_anomaly = signif(temp_anomaly,3))
# geo styling
g <- list(
scope = 'north america',
showland = TRUE,
landcolor = toRGB("grey83"),
subunitcolor = toRGB("white"),
countrycolor = toRGB("white"),
showlakes = TRUE,
lakecolor = toRGB("white"),
showsubunits = TRUE,
showcountries = FALSE,
resolution = 50,
projection = list(
type = 'azimuthal equal area',
rotation = list(lon = -100)
),
lonaxis = list(
showgrid = TRUE,
gridwidth = 0.5,
range = c(-100, -40),
dtick = 5
),
lataxis = list(
showgrid = TRUE,
gridwidth = 0.5,
range = c(35, 55),
dtick = 5
)
)
#dropdown
#zones<-df %>% dplyr::pull(mgmt_zone) %>% unique()
all_buttons <- list(
list(
label = "MAB",
method = "restyle",
args = list(
list(visible = c(TRUE,FALSE,FALSE)),
list(markers = list(
list(
visible = df %>% dplyr::filter(mgmt_zone == "MAB") %>%
dplyr::pull(temp_anomaly)),
c(),c())))),
list(
label = "NEC",
method = "restyle",
args = list(
list(visible = c(FALSE,TRUE,FALSE)),
list(markers = list(c(),
list(
visible = df %>% dplyr::filter(mgmt_zone == "NEC") %>%
dplyr::pull(temp_anomaly)),
c())))),
list(
label = "NED",
method = "restyle",
args = list(
list(visible = c(FALSE,FALSE,TRUE)),
list(markers = list(c(),c(),
list(
visible = df %>% dplyr::filter(mgmt_zone == "NED") %>%
dplyr::pull(temp_anomaly))
)))))
#sliders
udates<-df$yearmon %>% unique()
all_dates <- list()
for (i in 1:length(udates)) {
all_dates[[i]] <- list(
label = udates[i],
method = "restyle",
args = list(
list(visible = df %>% dplyr::filter(yearmon == udates[i]))
)
)
}
# annotations
annot <- list(x = 0, y=0.8, text = "Management\nZone", yanchor = 'bottom',
xref = 'paper', xanchor = 'right',
showarrow = FALSE)
# original interactive map
fig<- plot_geo(df)
fig <- fig %>% group_by(mgmt_zone)
fig <- fig %>% add_markers(x = ~lon, y = ~lat,
color = ~temp_anomaly, colors = turbo(99),
text = ~paste(df$temp_anomaly, "°C"), hoverinfo = "text") %>%
colorbar(title = "SSTa (°C)")
fig <- fig %>% layout(
showlegend = FALSE, geo = g
)
# plot with custom events
fig <- fig %>% layout(annotations = annot,
updatemenus = list(list(active = 0, x = 0, y = 0.8,
buttons=all_buttons)),
sliders = list(
list(
active = udates,
currentvalue = list(prefix = "Dates: "),
steps = all_dates)
)
)
fig# fig <- plot_geo(df, lat = ~lat, lon = ~lon, color = ~temp_anomaly,
# colors = turbo(99))
# fig <- fig %>% add_markers(
# text = ~paste(df$temp_anomaly, "°C"), hoverinfo = "text") %>%
# colorbar(title = "SSTa (°C)")
# fig <- fig %>% layout(title = '', geo = g)
#
# figdf<-NWA_flavors %>% filter(yearmon == as.yearmon("Sep 2015")) %>%
rename("mgmt_zone"="ET_ID") %>%
mutate(temp_anomaly = signif(temp_anomaly,3))
subplot(
plot_geo(df,x = ~lon, y = ~lat,
color = ~temp_anomaly, colors = turbo(99)) %>%
add_markers(text = ~paste(df$temp_anomaly, "°C"), hoverinfo = "text") %>%
colorbar(title = "SSTa (°C)") %>%
layout(showlegend = FALSE, geo = g),
plot_geo(VDM_Sep_2015,x = ~lon, y = ~lat,
color = ~suitability, colors = mako(99)) %>%
add_markers(text = ~paste(VDM_Sep_2015$suitability), hoverinfo = "text") %>%
colorbar(title = "Fishing Suitability") %>%
layout(showlegend = FALSE, geo = g)
)fig<- plot_geo(df)
fig <- fig %>% group_by(mgmt_zone)
fig <- fig %>% add_markers(x = ~lon, y = ~lat,
color = ~temp_anomaly, colors = turbo(99),
text = ~paste(df$temp_anomaly, "°C"), hoverinfo = "text") %>%
colorbar(title = "SSTa (°C)")
fig <- fig %>% layout(
showlegend = FALSE, geo = g
)Shifts in Spatial Distribution
Thinking this would be a slider again but instead of two plots it will be SST anomaly and then vessel suitability center of gravity for each management zone. Pretty much what we have for the CCS Futures